1. Field of the Invention
The invention is related to an ultrasonic flow measuring technology, and particularly, to an ultrasonic multi-channel flow measuring method for enhancing the accuracy of the flow or flow rate measurement, if pairs of ultrasonic transducers are mounted on a pipe that had been already arranged on a place.
2. Description of the Background
A general ultrasonic flow measuring method in a pipe has something in common as follows; a flow rate is obtained by measuring a flow velocity on a diametric line or a plurality of chords of a fluid flow section using an ultrasonic wave and multiplying the flow velocity by a fluid section area. For example, if the flow velocity VD is measured on the diametric line based on an ultrasonic one-channel flow measuring method using an ultrasonic wave, the flow rate Q is calculated as follows:
Q=Kxc2x7VDxc2x7Sxe2x80x83xe2x80x83(1)
Wherein, K less than 1.0 is a flow coefficient and S is a fluid flow section.
The flow measuring method is available under the condition that K=constant. In other words, a straight part of a pipe must be sufficiently longer. And, it is also preferable that Reynolds number greater than 104.
On the other hand, even through the flow velocity distribution is irregular and K is not constant, an ultrasonic multi-channel flow measuring method is known as a method capable of measuring a flow rate in a high accuracy.
Typical ultrasonic multi-channel flow measuring methods and apparatuses therefor are disclosed as follows:
U.S. Pat. No. 5,531,124 granted on Jul. 2, 1996
U.S. Pat. No. 4,646,575 granted on Jul. 25, 1987
Japanese Patent No. 2676321 granted on Jul. 25, 1997
The ultrasonic multi-channel flow measuring method has features as follows; as shown in FIG. 1, the flow rate Q is calculated in a manner to measure a flow velocity on a plurality of chords in parallel to a diametric line of a fluid section to reproduce a flow velocity distribution curve, calculate an average flow velocity VS of the fluid section and then multiply the average flow velocity VS by the fluid section area S, or in a manner to double-integrate the section area that is changed according to the flow velocity distribution and the diameter. Therefore, the ultrasonic multi-channel flow measuring method doesn""t need the flow coefficient.
Q=VSxc2x7Sxe2x80x83xe2x80x83(2)
or
Q=∫S∫V(r)S(r)drxc2x7drxe2x80x83xe2x80x83(3)
For it, even if the straight portion of the pipe is relatively shorter and the flow velocity distribution is asymmetrical, the flow rate can be somewhat accurately measured. As described above, the common point of the ultrasonic flow measuring methods exists in that the flow velocity is measured using an ultrasonic wave and multiplied by the section area S to calculate the flow rate.
The ultrasonic flow meter has a greatest feature as follows; unlike another flow meter, transducers for measuring the flow rate can be mounted on a pipe that has been already arranged. In other words, in case of the ultrasonic multi-channel flow meter, the flow rate measuring can be established only by the mounting of paired transducers mounted, even through the pipe has been previously arranged. Therefore, there has been developed a technology for mounting the paired transducers on the pipe without stopping the flow of water. Especially, even in case of a pipe of a greater inner diameter, it allows a flow measuring apparatus to be installed in-site after the completing of the pipe arrangement. It means that the manufacturing of a pipe portion of a flow meter having a larger volume and weight and its transportation into a working site is not necessary. The previous welding work of a flange for mounting the flow meter also is not required.
The ultrasonic flow meter to be mounted on the previously arranged pipe pays attention to the follows; in case of the ultrasonic multi-channel flow meter, a flow measuring error can be checked in a higher reliability in site. Because the flow velocity is measured on a plurality of chords using an ultrasonic wave, a flow velocity distribution curve is to be written out. Therefore, according to a shape of the flow velocity distribution curve, a computation programming error xcex4sw of a section average flow velocity VS can be checked.
And, if the flow velocity is measured on the chord by a transit time difference method, ultrasonic transit times t1 and t2, measuring errors xcex4t1 and xcex4t2, an error xcex4xcex94t of a time difference xcex94t=t2xe2x88x92t1, a measuring error xcex4L of a transit distance L, an error of a projective distance d=Lcos xcexa8 on L, etc. are checked thereby to confirm a flow velocity measuring error xcex4V. Like this, a flow measuring error xcex4Q can be indirectly measured, and a maximum flow measuring error xcex4Qmax that can be expected is as follows:
xcex4Qmax=xcex4V+xcex4SW+xcex4Sxe2x80x83xe2x80x83(4)
Wherein, xcex4S is a fluid section measuring error, and xcex4SW is an error of a section average flow velocity or a double-integration of Vs and S. A transit time difference method for measuring the flow velocity well-known is as follows:                     V        =                                            L              2                                      2              ⁢              d                                ⁢          Δ          ⁢                      xe2x80x83                    ⁢                      t                                          t                2                            -                              t                1                                                                        (        5        )            
Therefore, xcex4V is as follows:
xcex4V=(2xcex4L+xcex4d)+{square root over (xcex42t1+xcex42t2+xcex42xcex94t)}=(2xcex4L+xcex4d)+{square root over (2xcex42t1,2+xcex42xcex94t)}xe2x80x83xe2x80x83(6)
Wherein, xcex4L and xcex4d are a measuring error of interval distances L and d to be inputted into a flow velocity arithmetic logic processor or microprocessor. The symbols of L and d are not changed during the measuring of the flow velocity. But, errors xcex4L and xcex4d of the transit time measuring are represented as an average square error because an accidental error component is larger.
The computation programming error xcex4sw can be obtained using a computer for calculating the expressions (2) and (3) according to various flow velocity distribution curve. Therefore, the measuring errors xcex4V and xcex4SW can be checked in a higher reliability, but if the ultrasonic flow meter is mounted in site, it is very hard to check the fluid section measuring error xcex4S, exactly. For it, xcex4S becomes larger, so the flow rate measuring error can be increased. The cause is as follows:
If the fluid section is an ideal circular, its section S is as follows:                     S        =                              π            4                    xc3x97                      D            2                                              (        7        )            
Wherein, D is an inner diameter of a pipe.
But, it is not possible to measure the inner diameter of the pipe that has been already arranged, directly. A simplest method is to calculate the section area using the inner diameter of a pipe presented by a manufacturer, but the inner diameter might be different from that of the pipe already arranged. For it, it is not possible to confirm the fluid section measuring error xcex4S of the section area S. Furthermore, if a corrosive resistance layer is formed on the inner surface of the pipe, its thickness can""t be measured, exactly. The section of the pipe is not explicitly circular, because the pipe may be deformed in the process of the storage, transportation and its arrangement work, resulted from being oval. Under the environmental condition, a measuring error xcex4D of an inner diameter D may be largely increased. The fluid section measuring error xcex4S is as follows;
xcex4S=2xcex4Dxe2x80x83xe2x80x83(8)
If D=600 mm and its absolute error xcex94D=8 mm,       δ    D    =                    (                  8          600                )            xc3x97      100        ≈          1.34      ⁢      %      
The fluid section measuring error xcex4S is as follows:
xcex4S=2xc3x971.34=2.68%
Therefore, even through the flow velocity is exactly measured, the flow rate measuring error is not reduced to less than xcex4S. Particularly, in case of a pipe having a larger diameter, its section easily becomes a larger oval ness. As a result, if the ultrasonic transducer is mounted on the pipe that has already been arranged in order to measure the flow rate, it is difficult to exactly measure the fluid section area S. Further, the fluid section measuring error xcex4S becomes larger, and so the flow rate measuring error xcex4Q is increased.
When the flow rate Q is calculated by the double-integration of the expression (3) based on the ultrasonic multi-channel flow rate measuring method, assuming that the section S is circular, the function S(r) is used (r is a radius variable). If the section S is oval, not circular, the double-integration also has a larger error. As a flow velocity distribution curve V(r) is considered to be between intervals xe2x88x92R and +R as shown in FIG. 1, the double-integration error occurs due to a measuring error xcex4D of D=2R.
An object of the invention is to provide an ultrasonic multi-channel flow rate measuring method for significantly reducing the flow rate measuring error under the condition that the inner section area of a pipe can""t be measured, when a plurality of paired transducers are mounted on the pipe already arranged in site.
According to the invention, an ultrasonic flow rate measuring method comprises steps of measuring an inner diameter of a pipe in a manner to subtract a corrosive resistance layer and its possible thickness deviation from the maximum deviation of an inner diameter presented as a pipe size within the range of being able to be estimated, measuring a flow rate QI of the ideal circular section having the measured inner diameter based on the ultrasonic multi-channel flow rate measuring method, calculating a flow rate QII of the remaining section according to a flow velocity distribution curve and adding the flow rate QII to the flow rate QI to calculate a total flow rate Q. Therefore, the method can measure not only the flow rate in the ideally circular fluid section of the inner diameter of the pipe, but also the flow rate in a higher reliability even under the condition that the inner diameter of a pipe can""t be measure, exactly.
More concretely, when paired transducers are intended to mount on a pipe already arranged and realize an ultrasonic multi-channel flow rate measuring method, a predictable maximum deviation xcex94D of an inner diameter presented as a pipe standard, a thickness a of a corrosive resistance layer and its possible thickness deviation xcex94a are obtainable as follows:
D=DSxe2x88x92(xcex94D+2a+2xcex94a)xe2x80x83xe2x80x83(9)
The flow rate QI is measured with such like ideal circular section       S    1    =            π      4        ·          D      2      
having the inner diameter D according to the ultrasonic multi-channel flow rate measuring method, and then the remaining section SII is as follows:                               S          II                =                                            π              4                        ⁢                                          (                                                      D                    S                                    -                  D                                )                            2                                =                                    π              ⁡                              (                                                      R                    S                                    -                  R                                )                                      2                                              (        10        )            
The flow rate QII is calculated corresponding to the remaining section SII according to the flow velocity distribution curve and added to the flow rate QI in order to measure the total flow rate Q. The total flow rate Q is as follows:
Q=QI+QIIxe2x80x83xe2x80x83(11)
In that case, assuming that an error of QII is xcex4QII, xcex4QII takes an effect on the total flow rate Q as follows:                               δ          IIQ                =                                                            Q                I                            +                                                Q                  II                                ⁡                                  (                                      1                    +                                          δ                      QII                                                        )                                                                                    Q                I                            +                              Q                II                                              -          1          +                                    δ              QII                                      1              +                                                Q                  I                                                  Q                  II                                                                                        (        12        )            
If QI≈50QII, an increasing rate xcex4Q of the total flow rate measuring error depended upon xcex4QII is             δ      QII        51    .
Even if the increasing rate xcex4Q is over xcex4QII=20%, xcex4Q≈0.4%. Therefore, the flow rate Q can be measured in a higher accuracy even under the condition that the inner diameter is exactly not measured. Herein, it is noted that QI is exactly measured by using an ultrasonic wave in a fluid section of an ideal circle having an inner diameter D.